Spatial Data Visualization

CA Oil Spill Analysis

Analysis of CA Oil Spill Incidents

Overview

This script uses R to conduct spatial data visualization of 2008 California oil spill incidents.

Two data layers are combined, a California county layer and a CA DFW oil spill incident tracking dataset. The oil spill dataset is a vector layer.

Data citations: CA County layer dataset is provided in ESM244.

Cal EMA and DFG-OSPR (2009). Oil Spill Incident Tracking [ds394]. https://map.dfg.ca.gov/metadata/ds0394.html

hide

# Read in data and check/update projections
# Read in the CA county data (TIGER shapefile):
ca_counties_sf <- read_sf(here("data/counties"), layer = "CA_Counties_TIGER2016") %>% 
  janitor::clean_names() %>% 
  select(name)

# Check the projection
# st_crs(ca_counties_sf) # WGS 84 / Pseudo-Mercator, EPSG: 3857

# Read in the oil spill incident layer: 
oil_spill_sf <- read_sf(here("data/oil_spill_data"), layer = "ds394") %>% 
  janitor::clean_names()

# Check the projection:
# st_crs(oil_spill_sf) # NAD83 / California Albers, EPSG: 3310

# Make both have the same projection = NAD83 / California Albers, EPSG: 3310
ca_counties_sf <- st_transform(ca_counties_sf, st_crs(oil_spill_sf))

hide

### Plot the data for quick visualization
# Make a quick ggplot:
# ggplot() +
#   geom_sf(data = ca_counties_sf) +
#   geom_sf(data = oil_spill_sf)

Interactive map in tmap showing the location of oil spill events

hide

tmap_mode("view") # Allows one to drag the map and zoom in and out.

tm_shape(oil_spill_sf) +
  tm_dots("red")

Figure 1. Interactive overview of datapoints showing oil spill incidents in 2008.

Static choropleth map: fill color for each county depends on the count of inland oil spill events by county for the 2008 oil spill data

hide

# Join county dataset with oil dataset
county_oil_sf <- ca_counties_sf %>% 
  st_join(oil_spill_sf)

#head(county_oil_sf)

# Find counts of incidents per county
oil_counts_sf <- county_oil_sf %>% 
  filter(inlandmari == "Inland") %>% 
  group_by(name) %>%
  summarize(number = n()) # Do not want to just drop NA or else the county will disappear too - locations where id = NA will be replaced with 0. Count() would count every row, even those with NAs, so instead I will use group_by() and summarize().

# head(oil_counts_sf)

# Choropleth map
ggplot(data = oil_counts_sf) + # pull data from oil_counts_sf
  geom_sf(aes(fill = number), 
          color = "black", 
          size = 0.1) +
  scale_fill_gradientn(colors = c("lightgray","orange","red")) + # set colors
  theme_minimal() + # sets overall graph appearance
  labs(fill = "Number of Oil Spill Incidents") # sets title of legend

Figure 2. Choropleth map showing the count of inland oil spill incidents in 2008.